NY-ESO-1 is one of the most immunogenic cancer antigens known to date, inducing humoral and cellular immune responses in a high proportion of patients with advanced NY-ESO-1-expressing cancers. The assessment of spontaneous and vaccine-induced CD8+ T cell responses has been limited to a small number of known NY-ESO-1 epitopes presented by MHC class I alleles. Recently, a new method to monitor NY-ESO-1-specific CD8+ T cell responses was introduced that does not depend on the individual MHC class I status and on predefined peptide epitopes. Antigen-presenting cells transduced with recombinant adenoviral vectors encoding NY-ESO-1 were used to stimulate CD8+ selected NY-ESO-1-specific T cells. Effector cells were tested for recognition of autologous B cell targets transfected with NY-ESO-1 using a recombinant vaccinia virus construct. Using a modified approach we identified the NY-ESO-1 p94-102 peptide as being recognized by CD8+ T cells in the context of HLA- B51. NY-ESO-1 p94-102 specific CD8+ T cells recognized naturally processed NY-ESO-1 presented by HLA-B51+ monocyte-derived dendritic and tumor cells. Transfection of target cells with NY-ESO-1 combined with different HLA class I alleles confirmed that the NY-ESO-1 peptide was naturally processed and recognized by HLA-B51-restricted CD8+ T cell lines and clones. Therefore, NY-ESO-1 p94-102 is a new candidate peptide antigen for cancer immunotherapy and for the monitoring of spontaneous and vaccine-induced NY-ESO-1-specific T cell responses in HLA- B51+ patients with NY-ESO-1 expressing malignancies.
This article was published in Cancer Immunity, a Cancer Research Institute journal that ceased publication in 2013 and is now provided online in association with Cancer Immunology Research.
NY-ESO-1 was identified by serological analysis of a recombinant cDNA expression library (SEREX) from a squamous cell cancer of the esophagus (1). NY-ESO-1 belongs to the `cancer/testis´ family of antigens (2). NY-ESO-1 expression is observed in different types of cancer, including hematological malignancies, and is restricted in normal tissues to germ cells (3, 4, 5, 6, 7, 8). NY-ESO-1 is one of the most immunogenic tumor antigens known to date, eliciting spontaneous humoral and cellular immune responses in approximately 50% of patients with advanced NY-ESO-1 expressing cancers (9, 10). Spontaneous immunity against NY-ESO-1 is dependent on the presence of NY-ESO-1+ disease (10). Increases or decreases in NY-ESO-1 serum antibody titers and frequencies of NY-ESO-1-specific CD8+ T cell responses strongly correlate with progression or regression of NY-ESO-1+ cancers (10, 11).
A number of sensitive methods have been developed to monitor spontaneous and vaccine-induced immune responses against NY-ESO-1 in cancer patients. NY-ESO-1 serum antibodies, as assessed by ELISA or Western blot techniques, were identified as reliable indicators for CD8+ T cell responses (9, 10). The latter are detected by complementary methods, such as HLA-A2/peptide tetramers, ELISPOT and cytotoxicity assays (11, 12).
Studies of spontaneous and vaccine-induced T cell reactivity against NY-ESO-1 are restricted to the limited number of NY-ESO-1 epitopes known to be recognized by CD8+ T cells. Three NY-ESO-1 peptides that are recognized in the context of HLA-A2 have been described (13). Additional NY-ESO-1 epitopes presented by HLA-A31 were defined that are encoded by the normal and the alternative open reading frame of NY-ESO-1 (14). To broaden the patient population that can be evaluated for NY-ESO-1-specific T cell immunity, a new analytical approach was introduced by Gnjatic et al. which detects NY-ESO-1-specific CD8+ T cell reactivity independently of the predefined peptide epitopes presented by individual HLA class alleles. Antigen-presenting cells were transfected with adenovirus constructs encoding NY-ESO-1 to stimulate CD8+ selected effector cells. Autologous B cells were infected with vaccinia vectors encoding NY-ESO-1 and used as targets for specific CD8+ T cell recognition. The analysis of candidate NY-ESO-1 peptides presented by the respective HLA restriction elements has led to the identification of additional NY-ESO-1 epitopes presented by HLA-Cw3 and HLA-Cw6 (12). Applying this method, we identified a new NY-ESO-1 epitope that is recognized by CD8+ T cells from patients with detectable NY-ESO-1 serum antibody in the context of HLA-B51. CD8+ T cells recognized NY-ESO-1+, HLA-B51+ tumor cells presenting the naturally processed NY-ESO-1 p94-102 peptide. Therefore, the new NY-ESO-1 epitope is considered to be suitable for vaccination studies and the monitoring of NY-ESO-1-specific immune responses in HLA- B51+ patients with NY-ESO-1-expressing cancers.
Based on our previous study showing a correlation between the humoral and cellular response against NY-ESO-1, five patients with NY-ESO-1-expressing tumors and who were seropositive for NY-ESO-1 were selected to determine the nature of their CD8+ T cell responses.
Sensitization with recombinant adenovirus encoding NY-ESO-1 induces specific CD8+ T cell responses against NY-ESO-1 peptides
CD8+ selected effector T cells were sensitized with NY-ESO-1 recombinant adenovirus (Ad2/ESO)-infected, CD8-depleted PBLs of melanoma patient NW1539 and tested for the recognition of NY-ESO-1 epitopes on T2.A3 cells pulsed with long, overlapping 18-mer NY-ESO-1 peptides in ELISPOT assays. NY-ESO-1 peptides p85-102 and p91-108 were recognized (Figure 1). Since HLA-A3 and B51 were the only MHC class I alleles shared by the antigen-presenting T2.A3 cell line and the CD8+ effector T cells from patient NW1539, six candidate 9-mer peptides which can potentially bind to HLA-A3 or B51 and which are located within the 18-mer sequences recognized were synthesized and tested for recognition by Ad2/ESO-presensitized CD8+ NW1539 T cells. Figure 2 shows the results of an ELISPOT assay against T2.A3 target cells pulsed with NY-ESO-1 9-mer peptides recognized by CD8+ NW1539 T cells. A strong reactivity was demonstrated against NY-ESO-1 p94-102 (Figure 2). The specific reactivity against NY-ESO-1 p94-102 was confirmed by the analysis of peptide-presensitized CD8+ T cells derived from four patients who were seropositive for NY-ESO-1 and who share the HLA-A3 and/or the B51 allele (Figure 3).
Specific recognition of NY-ESO-1 p94-102 in the context of HLA-B51
The CD8+ T cell clone NW1539-CTL-1/15, obtained by repeated in vitro stimulations with the autologous tumor cell line NW-MEL-1539, was tested against NY-ESO-1 p94-102 peptide-pulsed T2 and T2.A3 target cells to discriminate between HLA-A3 and B51 as the restriction element. As shown in Figure 4, both peptide-pulsed T2 and T2.A3 cells were recognized with comparable efficacy, suggesting that NY-ESO-1 p94-102 recognition was not dependent on the presence of HLA-A3, leaving HLA-B51 as the potential restriction element for p94-102 presentation.
To confirm that NY-ESO-1 p94-102 was presented by HLA-B51, we tested peptide-pulsed PHA blasts derived from different donors as targets for recognition by CD8+ T cells from melanoma patient NW1274 presensitized with NY-ESO-1 p94-102 peptide. Figure 5 shows a strong recognition of NY-ESO-1 p94-102 peptide-pulsed HLA-B51+ target cells NW1725 and NW1274. No CD8+ T cell reactivity was found against HLA-B51 negative target cells MZ7 and NW1726.
Recognition of naturally processed NY-ESO-1 by peptide-sensitized CD8+ effector T cells
CD8+ selected effector T cells of the HLA-B51+ melanoma patient NW1539 were presensitized with NY-ESO-1 p94-102 and tested in ELISPOT assays on autologous, monocyte-derived APCs pulsed with the stimulating peptide or infected with Ad2/ESO. Figure 6 shows the specific recognition of NY-ESO-1 p94-102 and Ad2/ESO-transfected APCs, suggesting that NY-ESO-1 peptide p94-102 is naturally processed. APCs pulsed with the NY-ESO-1 recombinant protein, the SSX protein, or a lysate of the NY-ESO-1 expressing tumor cell line NW-MEL-38 were not recognized.
Recognition of naturally processed NY-ESO-1 is restricted to HLA-B51
To ascertain the HLA class I restriction element for NY-ESO-1 p94-102 peptide-specific CD8+ T cell responses against naturally processed NY-ESO-1, we tested COS-7 cells transfected with NY-ESO-1 and either HLA-A3 or B51. Figure 7 shows that the HLA-A3+, B51+ effector T cell lines NW923 and NW1539 specifically recognize NY-ESO-1 in the context of HLA-B51, and not of HLA-A3. These results suggest that the NY-ESO-1 p94-102 epitope is naturally processed and presented only by HLA-B51.
Specific cytotoxicity of T cell clone NW1539-CTL-1/1 against NY-ESO-1 p94-102 and naturally processed NY-ESO-1
T cell clone NW1539-CTL-1/1 was generated by limiting dilution and repeated in vitro stimulation with NY-ESO-1 p94-102 peptide. Cytotoxicity of T cell clone NW1539-CTL-1/1 against NY-ESO-1 p94-102 peptide-pulsed T2.A3 cells, autologous vvESO-transduced NW1539-EBV B cells, the NW1539-MEL-1 melanoma cell line and the allogeneic LB39 melanoma cell line, which shares only the HLA-B51 allele with the NW1539-CTL-1/1 effector cells, is demonstrated in Figure 8. No lysis was observed against untreated T2.A3, K562, and NW1539-EBV B cells.
Spontaneous ex vivo reactivity of CD8+ NW1539 T cells against NY-ESO-1 p94-102
Unsensitized CD8+ selected T cells from melanoma patient NW1539 were analyzed for specific recognition of NY-ESO-1 p94-102 in ELISPOT assays. The high number of peptide-specific spots shown in Figure 9 suggests a precursor frequency of NY-ESO-1 p94-102-reactive T cells of approximately 0.7% of the CD8+ T cell population in this patient. Rare cases of comparable ex vivo precursor frequencies against Melan A/MART-1 p26-34 (15, 16) and NY-ESO-1 p157-167 (17) were reported for individual patients during prolonged vaccination with the respective peptides. In this light, the spontaneous CD8+ T cell reactivity against NY-ESO-1 p94-102 supports previous findings pointing to the strong immunogenicity of NY-ESO-1 in cancer patients.
The cancer/testis antigen NY-ESO-1 is one of the most immunogenic tumor antigens known to date. NY-ESO-1 serum antibodies are detected in half of the patients with advanced NY-ESO-1 expressing malignancies (9, 11). NY-ESO-1-specific CD8+ T cell responses assessed in a population of HLA-A2+ patients with advanced NY-ESO-1-expressing cancers showed a close correlation with detectable NY-ESO-1 serum antibody (11, 12). To extend the monitoring of T cell immunity against NY-ESO-1 to epitopes presented by non-HLA-A2 alleles, Gnjatic et al. have introduced a new method that circumvents the need for established tumor and T cell lines, as well as for predefined peptide epitopes presented by respective HLA restriction elements. Antigen-presenting cells transfected with NY-ESO-1 recombinant adenoviral constructs were used to stimulate CD8+ effector T cells. These effectors were tested for the recognition of autologous or HLA-matched allogeneic NY-ESO-1 vaccinia virus transduced EBV-B cells. Two naturally processed NY-ESO-1 peptides were identified with this approach that are presented by HLA-Cw*03 (NY-ESO-1 p92-100) and HLA-Cw*06 (NY-ESO-1 p80-88) (12).
To expand the list of MHC class I-restricted NY-ESO-1 peptides that can be used to monitor NY-ESO-1-specific T cell responses and to vaccinate cancer patients, we selected NY-ESO-1 seropositive patients sharing the HLA-A3 and B51 allele for the analysis of spontaneous NY-ESO-1-specific CD8+ T cell responses. Ad2/ESO- or NY-ESO-1 p94-102 peptide-sensitized CD8+ selected effector cells of four patients (NW1539, NW1354, NW923, NW1352) with NY-ESO-1+ cancers specifically recognized HLA-A3+ and B51+ T2.A3 cells pulsed with long overlapping 18-mer NY-ESO-1 peptides (NY-ESO-1 p85-102 and p91-108). The long NY-ESO-1 peptides recognized were analyzed for potential binding motifs to HLA-A3 and/or B51. Six NY-ESO-1 9- and 10-mer peptides (Table 1) were tested and peptide p94-102 was recognized by Ad2/ESO-sensitized NW1539 T cells. Analysis of four additional HLA-A3 and/or B51+ patients confirmed the recognition of NY-ESO-1 p94-102 peptide.
The NY-ESO-1-specific CD8+ T cell clone NW1539-CTL-1/15 was found to recognize NY-ESO-1 p94-102 peptide-pulsed onto T2 and T2.A3 cells. HLA-B51 is the HLA class I allele shared by both target and effector cells, suggesting that NY-ESO-1 p94-102 is presented by and recognized in the context of HLA-B51. HLA-B51 restriction was confirmed by testing peptide-pulsed target cells expressing different HLA class I alleles for their ability to recognize peptide-presensitized CD8+ T cells from HLA-A3-, B51+ melanoma patient NW1274.
Antigenic peptides derived from tumor antigens may be considered for the vaccination of cancer patients if they represent naturally processed epitopes that can be recognized by CD8+ T cells when presented on antigen expressing tumor cells (18, 19, 20, 21, 22, 23, 24, 25). Circumventing the difficulty of establishing tumor cell lines with different tumor antigen- and HLA profiles, the gene of interest can be expressed in monocyte-derived antigen-presenting cells or EBV-B cells and tested for the recognition of naturally processed epitopes by HLA-matched CD8+ effector T cells. Co-transfection experiments with NY-ESO-1 along with HLA-A3 or B51 showed that natural processing of NY-ESO-1 p94-102 was restricted to HLA-B51. NY-ESO-1 p94-102 peptide-presensitized CD8+ T cells recognized NY-ESO-1-transduced, autologous, monocyte-derived APCs and the HLA-B51+, NY-ESO-1+ LB39 melanoma cell line, suggesting that NY-ESO-1 p94-102 represents a naturally processed epitope. Recognition of NY-ESO-1 p94-102 peptide by effector T cells derived from HLA-B51 negative individuals may result from cross-presentation of the antigen in the context of an additional HLA class I molecules yet to be defined.
We have recently evaluated the immunological and clinical effects of intradermal immunization with the HLA-A2-restricted NY-ESO-1 epitopes p157-167 (SLLMWITQCFL), p157-165 (SLLMWITQC), and p155-163 (QLSLLMWIT) in a clinical trial (22). Strong inflammatory delayed-type hypersensitivity reactions associated with primary CD8+ T cell responses against NY-ESO-1 p157-167 and p157-165 were induced in the majority of patients without spontaneous immunity to NY-ESO-1. Stabilization and regression of metastatic disease observed in individual patients coincided with the induction of peptide-specific CD8+ T cell responses in vivo (22). The potential role of the new NY-ESO-1 p94-102 epitope in clinical vaccine trials will be assessed using recombinant HLA-B51/peptide multimers, ELISPOT and cytotoxicity assays to determine the immunogenicity of this peptide in a larger series of NY-ESO-1 seropositive and seronegative, HLA-B51+ cancer patients. If the correlation of spontaneous NY-ESO-1 serum antibody with detectable CD8+ T cell reactivity can be confirmed in HLA-B51+ cancer patients with the new NY-ESO-1 p94-102 epitope, the monitoring of spontaneous and vaccine-induced immune responses against NY-ESO-1 could be extended to HLA-B51+ patients.
Materials and methods
NY-ESO-1 serum antibody
Serum antibody responses against the recombinant NY-ESO-1 protein were tested by standard Western blot analysis and ELISA, using NY-ESO-1 recombinant protein purified from E. coli as described (9).
Five patients with advanced NY-ESO-1-expressing cancer and detectable NY-ESO-1 serum antibody were selected for the assessment of peptide-specific CD8+ T cell responses. Diagnoses and HLA class I types were as follows: Melanoma NW1539 (A*03, A*11; B*44, B*51; Cw*04, Cw*05); melanoma NW1352 (A*03, A*31; B*15, B*35; Cw*03, Cw*04); urothelial carcinoma NW923 (A*01, A*03; B*51, B*52; Cw*06, Cw*012); non-small cell lung cancer NW1354 (A*03, A*11; B*35, B*44; Cw*04, Cw*016); melanoma NW1274 (A*2; B*44, B*51; Cw*1-7 not determined). PHA blasts used as antigen-presenting target cells were prepared from peripheral blood lymphocytes (PBL) of two healthy donors, NW1725 (A*32; B*51; Cw*1-7 not determined) and NW1726 (A*2; B*7, Bw*6; Cw*7).
The tumor cell lines NW-MEL-38 (A*01, A*02; B and C locus not determined), NW-MEL-1539 (A*03, A*11; B*44, B*51; Cw*04, Cw*05), and LB39 (A*02, A*24; B*07, B*51; Cw*01, Cw*07), a generous gift from Dr. Thierry Boon, were cultured in Dulbecco's modified Eagle's medium (DMEM, GIBCO) containing 10 mM HEPES buffer, L-arginine (84 mg/l), L-glutamine (584 mg/l), penicillin (10 IU/ml), streptomycin (100 µg/ml), and 10% FCS. LB39 was treated with 100 U/ml IFN-gamma for 5 days to upregulate MHC class I before being used in the cytotoxicity assay. EBV-transformed B lymphocytes, NW115-EBV and MZ1257-EBV, were used as feeder cells for the T cell culture. The mutant cell lines CEMx721.174.T2 (T2) and CEMx721.174.T2 transfected with HLA-A3 (T2.A3), kindly provided by Dr. Vincenzo Cerundolo, were maintained in RPMI 1640 medium supplemented with 10 mM HEPES buffer, L-arginine (242 mg/l), L-asparagine (50 mg/l), L-glutamine (300 mg/l), penicillin (10 IU/ml), streptomycin (100 µg/ml), 1% non-essential amino acids and 10% FCS.
The CD8+ T cell line NW1539-IVS-1 and the T cell clone NW1539-CTL-1/15 were generated using PBLs from patient NW1539 and were used for in vitro stimulation of the autologous tumor cell line NW-MEL-1539. The CD8+ T cell clone NW1539-CTL-1/1 was obtained by limiting dilution and repeated in vitro stimulations with NY-ESO-1 p94-102.
Peptides 18 amino acids in length with overlapping sequences spanning the entire NY-ESO-1 sequence have been described previously (26). NY-ESO-1 nonamer and decamer peptides were synthesized by Multiple Peptide Systems (San Diego). The purity of the peptides was >95%. The sequences of these peptides are given in Table 1.
Adenoviral constructs Ad2/EGFP (encoding green fluorescent protein) and Ad2/ESO (encoding NY-ESO-1) were provided by Genzyme Corporation, Farmington MA. Vaccinia constructs vvWT (wild type vaccinia virus) and vvESO (encoding the full length NY-ESO-1 cDNA) were provided by Therion Biologics, Cambridge MA. These have been described previously (12).
Presensitization with peptides and adenoviral constructs
Purified CD8+ T lymphocytes were presensitized with peptide-pulsed or adenovirus-infected irradiated autologous PBLs (depleted of CD4+ and CD8+ T cells) as described. Presensitized CD8+ T cells were used as effectors on day 6 for ELISPOT analysis or restimulated on day 7 to assess cytotoxicity against peptide-pulsed T2.A3 cells, adenovirus-infected autologous APCs or melanoma cells in 51chromium release assays 7 days later.
To generate monocyte-derived, autologous antigen-presenting cells, peripheral blood mononuclear cells (PBMCs) depleted of CD4+ and CD8+ T lymphocytes using magnetic beads (MiniMACS, Miltenyi Biotec, Bergisch Gladbach, Germany) were seeded in 24-well plates at 4 x 106 cells/well and allowed to adhere to plastic for 24 h. Non-adherent cells were removed and the remaining cells were used as antigen-presenting cells (APCs). These were cultured with 1000 U/ml GM-CSF (Leukomax, Sandoz, Nürnberg, Germany) and 1000 U/ml IL-4 (Pharma Biotechnologie Hannover, Germany) for 5 days in 2 ml/well X-vivo 15 medium (Bio Whittaker, Walkersville, Maryland, USA). To assess the HLA-B51-restricted presentation of NY-ESO-1, APCs were treated on day 6 of in vitro culture with 1000 U/ml IL-4, 1000 U/ml IL-6, 10 ng/ml IL-1beta, 10 ng/ml TNF-alpha (all from Pharma Biotechnologie Hannover, Germany), 1000 U/ml GM-CSF, and 1 µg/ml prostaglandin (Sigma Chemical Co., St. Louis, MO). APCs were infected with adenoviral constructs at 1000 infectious units/cell or pulsed with peptides at 10 µg/ml and cultured for 24 h. The APCs were then washed twice and used as targets in ELIPOT assays at 3 x 104 cells/well.
PHA blasts generated from PBLs by incubation with 0.5 µg/ml PHA (Welcome) were pulsed with 10 µg/ml peptide. EBV-B cells were either pulsed with 10 µg/ml peptide or infected overnight with adeno- or vaccinia wild type or recombinant for NY-ESO-1 at 30 pfu/cell in 300 µl serum-free medium.
Effector cells, either CD8+ T lymphocytes presensitized as described above (2.5 x 104 cells/well) or T cell clones (1000 cells/well), were added to 96-well flat-bottom nitrocellulose plates (MAHA S45 10, Millipore, Bedford MA, USA) coated with 5 µg/ml anti-IFN-gamma antibody (Hölzel Diagnostic, Köln, Germany) in a final volume of 100 µl. Irradiated T2 and T2.A3 cells were pulsed with NY-ESO-1 peptides at a concentration of 10 µg/ml, washed and added to the effector cells at 5 x 104 cells/well. Plates were incubated for 16 h at 37°C, washed extensively (6 times with 0.05% Tween 20 in PBS) and 0.5 µg/ml biotinylated anti-IFN-gamma detection antibody (Hölzel Diagnostic, Köln, Germany) was added. Following incubation for 2 h at 37°C, the plates were washed and developed with ABC-AP Vectastain (Vector, Burlingame, CA, USA) for 1 h. Following the addition of substrate (BCIP/NPT) and a 10 min incubation, the plates were prepared for microscopy and the number of blue spots/well, corresponding to the proportion of NY-ESO-1-specific CD8+ T cells/well, were counted.
Cytotoxicity against peptide-pulsed T2.A3 cells, NY-ESO-1-transduced EBV-B cells, and tumor cell lines was determined in standard chromium release assays as described (11).
COS cell transfections
The full-length NY-ESO-1 sequence was isolated from the corresponding pQE9 clone and cloned into the BamHI-HindIII sites of the pcDNA3.1(-) vector (Invitrogen). Transfection of COS-7 cells was carried out as described (27). Briefly, 2 x 104 COS-7 cells were transfected with 150 ng pcDNA3.1(-) containing NY-ESO-1 cDNA and 150 ng pcDNA1Amp containing HLA-A3 or HLA-B51 cDNA using the DEAE-dextran-chloroquine method. The HLA-A3 and HLA-B51 cDNAs were kindly provided by Drs. Thomas Wölfel and Thierry Boon respectively. Transfectants were incubated at 37°C for 48 h and tested in a T cell stimulation assay after 24 h.
Immunoscreening of NY-ESO-1 transfectants
Transfectants were tested for their ability to stimulate TNF-alpha production by the CD8+ T cell lines NW923-IVS-1 and NW1274-IVS-1 (generated by repeated in vitro stimulation with NY-ESO-1 p94-102) and by NW1539-IVS-1 (generated by repeated in vitro stimulation with the autologous tumor cell line NW-MEL-1539) as described (27). Briefly, 5000 CD8+ T cells in 100 µl RPMI supplemented with 10% human serum and 25 U/ml rhuIL-2 were added to microwells containing COS-7 transfectants. After 24 h, 50 µl of supernatant were collected and the TNF-alpha content determined by testing the cytotoxicity against WEHI 164 clone 13 cells in an MTT colorimetric assay.
This work was supported by the Cancer Research Institute/Partridge Foundation, New York, U.S.A. and Krebsforschung Rhein Main, Frankfurt, Germany.
- Received August 5, 2002.
- Accepted August 21, 2002.
- Copyright © 2002 by Elke Jäger